Having improved my Thing-O-Matic mechanics about as much as can be cough reasonably expected, the stepper motors driving the X and Y axes still seem to be running at about the limit of their ability. It’s time for some doodling on that subject; let’s start by collecting all the data in one spot.
The X and Y motors are, as far as I can tell, inherited directly from the MBI Cupcake CNC. They seem to be Kysan 42BYG034-4.78 (aka SKU 1123029) described on that Kysan store product page as:
- 14V
- 0.4A
- 35 OHM
- 44MH
- 3/16 SHAFT
- 8 DEGREE (error: should be 1.8°)
- 2.3KG-CM MAX
- 20G.CM2
The Kysan Electronics product page has more data, including a torque curve:
Note that, unlike most NEMA 17 steppers, this puppy does not have a 5 mm shaft (unlike the electrically identical Kysan 42BYG034, which does). A 5 mm pulley is a poor fit on a 4.78 mm shaft and, conversely, you must drill / bore MBI pulleys to fit other steppers.
If one was to buy a replacement pulley with a 5 mm bore, the A 6D51M018DF0605 from SDP might do the trick. Or you could apply a 0.199 inch (#8) drill to the bore and save twenty bucks.
The Z stepper has an integrated 4-start leadscrew, so it’s not suitable for a drop-in replacement without some Quality Shop Time. That comment leads to that Kysan store product page for Kysan 17HD011-200N (aka SKU 1040104), with this data:
- 12V
- 0.4A
- 200MM LEAD SCREW
- 1.8 DEGREE
- 30 OHM
- 37MH
- 260mN.m HOLDING TORQUE
The Kysan Electronics product page has more data (although the proffered PDF datasheet is empty), including a low-res torque curve:
The Z axis motor on my TOM is labeled “PN 1040103”, which leads to that Kysan Electronics product page for 17HD-8X150MM0.4A, with just a mechanical drawing. The electrical properties seem identical.
The MK6 Extruder (MBI assembly and setup doc there) stepper motor evidently arrives without ID, but that comment suggests it’s an Anaheim Automation 17Y402S-LW4-01 or something similar. That discussion indicates the motor has a 5 mm shaft, not the 6 mm that would match the standard MK5/MK6 filament drive gear, and that MBI did another custom order.
Following various links leads us to the data table there, wherein we find:
- Coil resistance = 12 Ω
- Coil inductance = 29 mH
- Rated current = 850 mA
- Rated voltage = 10.2 V
The torque curve:
One observation: these motors have extremely high winding resistance. That makes them suitable for low-speed H-bridge drivers without current control, not contemporary stepper drivers with chopper current control.
The Smell of Molten Projects in the Morning 
I’ve been playing with the pic-based linistepper stepper motor driver:
http://www.piclist.com/techref/io/stepper/linistep/index.htm
They do a pretty good job for higher-inductance steppers, and aren’t too expensive. (Additional fan cooling pretty much mandatory, though.)
I have them running the three axes on my sherline and thus far they’ve been the least of my problems.
That, I think, falls in the category of “to be damned by faint praise”… [grin]
I’m not convinced that conventional microstepping drivers are all that bad. The power loss due to hysteresis and suchlike is less than the good old I2R dissipation from the winding current itself, sooo it’s not clear toasting the driver makes for a net improvement.
Stepper motors have a surprisingly high operating temperature limit, but even with that in mind I think some folks play fast-and-loose with the motor’s maximum power limit. Typical NEMA 17 motors have a few ohms resistance and a 1 A winding limit, so they dissipate a few watts, tops. Driving both windings (or all four windings) at their rated current overpowers the poor things; the motor power limit comes from a single winding at its rated voltage & current.
I’ll grant that linear step interpolation will be smoother, but it’s not clear (at least to me) it actually makes a difference over a well-tuned conventional driver in most applications.
But I’ve been wrong in public before…
>That, I think, falls in the category of “to be damned by faint praise”
I’m a deeply suspicious person: the linisteppers haven’t let me down *yet*… (unlike a lot of other things in this project.)
The only complaint I have with them is they use standard headers for all the input/output, and I’d rather have something latching/locking, but the newer rev of the board has already fixed that.
I’m in agreement with you that reducing power dissipation in the motor is probably a solution in search of a problem. I got them because they seem to have a pretty good price/performance ratio and at the time, probably four years ago, I didn’t have the knowledge to make my own. Now I think I’d build my own around the A3977’s (but I haven’t gotten time to see how they work in actual running conditions, so who knows?)
I just got a quartet of those impossibly small Pololu A4988 driver boards on the off chance that they’d turn out to be useful. They’re sufficiently awkward that Yet Another Carrier Board seems to be in the offing; I’d want all four mounted with more-or-less MBI compatible connectors, warm-and-fuzzy LEDs, and some heatsinking. The PCB layout looks pretty good, too.